Relative Position Matters: Trajectory Prediction and Planning with Polar Representation
Bozhou Zhang, Nan Song, Bingzhao Gao, Li Zhang
AI summary
Problem
Existing Cartesian-based methods implicitly model the varying influence of surrounding traffic elements on the ego vehicle, leading to suboptimal spatial relationship encoding.
Approach
The authors propose Polaris, a framework that encodes scene context and decodes trajectories entirely in polar coordinates, using a Relative Embedding Transformer to explicitly capture distance and directional relationships.
Key results
- State-of-the-art performance on Argoverse 2 trajectory prediction
- Superior open-loop and closed-loop planning scores on nuPlan
- Explicit polar coordinate encoding for agents and HD maps
- Relative Embedding Transformer for direct distance and angle modeling
Why it matters
Provides a more intuitive and accurate spatial representation for autonomous driving systems, benefiting researchers and engineers in trajectory prediction and planning.
Abstract
Trajectory prediction and planning in autonomous driving are highly challenging due to the complexity of pre- dicting surrounding agents’ movements and planning the ego agent’s actions in dynamic environments. Existing methods encode map and agent positions and decode future trajectories in Cartesian coordinates. However, modeling the relationships between the ego vehicle and surrounding traffic elements in Cartesian space can be suboptimal, as it does not naturally capture the varying influence of different elements based on their relative distances and directions. To address this limita- tion, we adopt the Polar coordinate system, where positions are represented by radius and angle. This representation provides a more intuitive and effective way to model spa- tial changes and relative relationships, especially in terms of distance and directional influence. Based on this insight, we propose Polaris, a novel method that operates entirely in Polar coordinates, distinguishing itself from conventional Cartesian- based approaches. By leveraging the Polar representation, this method explicitly models distance and direction variations and captures relative relationships through dedicated encoding and refinement modules, enabling more structured and spatially aware trajectory prediction and planning. Extensive experi- ments on the challenging prediction (Argoverse 2) and planning benchmarks (nuPlan) demonstrate that Polaris achieves state- of-the-art performance. We will release our code.